A dataset created in order to test, refine and explore a system of categorization of speech and thought presentation for English narrative texts, based on the one informally described in 'Style in Fiction', Leech and Short, 1981.